summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* BQ:ack batches of messagesAlexandru Scvortov2012-02-161-9/+9
|
* don't dead-letter on purgeAlexandru Scvortov2012-02-161-15/+1
|
* don't dead letter on queue deletionAlexandru Scvortov2012-02-161-43/+8
|
* rename ChPid to SenderPid where appropriateAlexandru Scvortov2012-02-151-13/+15
|
* dead-letter when auto-deletingAlexandru Scvortov2012-02-151-12/+24
|
* oopsAlexandru Scvortov2012-02-141-1/+1
|
* don't hang on multiple queue.purgesAlexandru Scvortov2012-02-141-21/+24
| | | | | | | Following Matthias' suggestion, we don't ok multiple deletes for the same queue: the first delete gets ok'd, but the other ones will get an exception. Because confirming multiple purges on the same queue is fine, but confirming multiple deletes on one queue is weird (if not wrong).
* merge default into bug20337Alexandru Scvortov2012-02-143-81/+63
|\
| * cosmeticMatthias Radestock2012-02-141-2/+2
| |
| * always issue credits, even when we drop a messageMatthias Radestock2012-02-141-30/+18
| | | | | | | | | | ...due to a dead-letter cycle and make the code a whole lot simpler
| * cosmeticMatthias Radestock2012-02-142-28/+24
| |
| * Merge headsEmile Joubert2012-02-141-9/+12
| |\
| | * Merge bug24714Simon MacMullen2012-02-141-12/+7
| | |\
| | * | monitor nodes just onceMatthias Radestock2012-02-091-9/+12
| | | |
| * | | Merge bug20337 into defaultEmile Joubert2012-02-141-12/+7
| |\ \ \ |/ / / / | | | _
| * | simplifying refactorMatthias Radestock2012-02-131-12/+7
| |/
* | factor out redundant codeAlexandru Scvortov2012-02-141-15/+13
| |
* | refactoringAlexandru Scvortov2012-02-142-20/+25
| |
* | don't hang if purging an empty queue with dlxAlexandru Scvortov2012-02-121-2/+6
| | | | | | | | The queue.delete case had already been handled.
* | use lists:keydelete and only log cycles onceAlexandru Scvortov2012-02-122-14/+14
| |
* | merge default into bug20337Alexandru Scvortov2012-02-1211-94/+470
|\ \ | |/ |/| | | All tests pass. This also fixes the rabbit_guid rename.
| * cosmeticAlexandru Scvortov2012-02-061-24/+21
| | | | | | | | The changes on default don't seem to impact the working of this branch.
| * merge default into bug20337Alexandru Scvortov2012-02-06131-657/+1040
| |\
| * | refactor (Matthias's suggestions)Alexandru Scvortov2012-02-031-6/+2
| | |
| * | drop messages in dl cyclesAlexandru Scvortov2012-02-031-2/+45
| | | | | | | | | | | | | | | | | | This is the fate of messages in cycles: =WARNING REPORT==== 3-Feb-2012::12:54:34 === Message dropped. Dead-letter queues cycle detected: [<<"foo">>,<<"foo">>]
| * | use same variable naming convention as in channelAlexandru Scvortov2012-01-311-20/+21
| | |
| * | don't leak queue monitorsAlexandru Scvortov2012-01-311-3/+41
| | |
| * | check that dlx is defined is dl-rk is setAlexandru Scvortov2012-01-301-2/+5
| | |
| * | check dlx routing key argAlexandru Scvortov2012-01-241-1/+11
| | |
| * | overwrite malformed x-death headersAlexandru Scvortov2012-01-241-3/+4
| | |
| * | don't discard accumulated changesAlexandru Scvortov2012-01-241-5/+5
| | |
| * | use POSIX timestamps instead of human readable onesAlexandru Scvortov2012-01-241-2/+2
| | |
| * | merge default into bug20337Alexandru Scvortov2012-01-2447-447/+596
| |\ \
| * | | add setting dlx routing key in queue.declareAlexandru Scvortov2011-12-052-9/+28
| | | |
| * | | record routing keys and exchanges when dead-letteringAlexandru Scvortov2011-12-052-7/+16
| | | | | | | | | | | | | | | | | | | | | | | | | | | | We don't record BCC routing keys, only basic.publish and CC ones. Also, there's little reason to prefix the individual fields in the "x-death" table with "x-death-", so don't.
| * | | include the time the message was dead-letteredAlexandru Scvortov2011-12-051-1/+3
| | | |
| * | | merge default into bug20337Alexandru Scvortov2011-11-2810-81/+328
| |\ \ \
| | * | | fix tuple arity mismatchAlexandru Scvortov2011-11-281-2/+2
| | | | | | | | | | | | | | | | | | | | I strongly dislike dynamically-typed languages.
| | * | | don't hang if there are no DLQs bound to the DLXAlexandru Scvortov2011-11-281-13/+24
| | | | |
| | * | | death headers are now stored in an array of tablesAlexandru Scvortov2011-11-281-8/+17
| | | | |
| | * | | don't check dlx on queue.declareAlexandru Scvortov2011-11-281-5/+3
| | | | |
| | * | | only remove messages when ALL DLQs have confirmedAlexandru Scvortov2011-11-151-48/+59
| | | | | | | | | | | | | | | | | | | | Also, simulate confirms when a DLQ dies.
| | * | | monitor the DLQsAlexandru Scvortov2011-11-151-8/+28
| | | | |
| | * | | removed reminder commentAlexandru Scvortov2011-11-151-2/+1
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | There seems to already be a mechanism in place for this. amqqueue_process:next_state/1 confirms the messages reported by VQ:drain_confirmed/1, and this seems to work fine for the other cases.
| | * | | wait for DLQ to confirm messages before deleting the queueAlexandru Scvortov2011-11-151-16/+23
| | | | |
| | * | | partial wait-for-DLQ-before-deleting-queueAlexandru Scvortov2011-11-141-50/+46
| | | | |
| | * | | spread out the dead-lettering on queue deletionAlexandru Scvortov2011-11-141-7/+8
| | | | |
| | * | | don't remove purged messages until the DLQ has confirmed themAlexandru Scvortov2011-11-141-27/+50
| | | | | | | | | | | | | | | | | | | | Also, block the queue.purge_ok until the messages have reached the DLQ safely.
| | * | | don't remove expired messages until the DLQ has confirmed themAlexandru Scvortov2011-11-142-7/+11
| | | | |
| | * | | don't remove rejected message before it is confirmed by the DLQAlexandru Scvortov2011-11-148-123/+130
| | | | |
View Lorry Controller Status